The Argument for Encouraging Student Use of AI in First-Year Law

A consensus is emerging in law schools across Canada and the US that AI has no place in teaching the core curriculum, because it impedes student learning, especially in first year. Schools are beginning to follow the University of Chicago’s lead in banning AI in first-year core courses. Some profs are also banning laptops in classrooms — taking every precaution to shield students’ developing minds from the corrosive effects of AI.

Michael Plaxton, one of the leading voices in Canada thinking through the many challenges AI poses to legal education, crystallizes the argument:

“The case for law schools must be premised on the value of imposing slowness and struggle on students. Otherwise, there is simply no way to explain why students should not acquire a bespoke, individualized legal ‘education’ by plugging prompts into Claude or ChatGPT.”

The Chicago policy similarly describes “the value of effortful struggle” during 1L as “paramount.”

The theory underlying this “make them struggle” approach is that students can build a sound grounding in law only by grinding through the 1L material. And any reliance on AI would prevent that.

Plaxton is explicit about this: using AI in 1L would lead to “never-skilling,” a term that, as he notes, comes from the medical literature to describe what can happen when medical students use AI in their clinical training. They learn less than they otherwise would.

I want to suggest — contrary to this emerging consensus — that what students go through in 1L of law school differs from other domains in which AI may cause “never-skilling.” Empirical evidence supports the opposite.

Student use of AI at the earliest stage of law school could help students learn more effectively and build a firmer foundation for first year and the rest of the degree.

The argument, briefly, is this: just as it helps to give students headnotes and guidance on the salient points in the cases early on, it would help to encourage 1Ls to create AI summaries and to dialogue with AI about cases to clarify points of confusion. This would demystify the material, make the law intelligible sooner, and help students grasp more of what 1L covers. It would also leave more time to work on higher-order skills like issue-spotting and formulating legal arguments.

Don’t underestimate how confusing 1L tends to be for students

The Juris Doctor degree in Canada and the US has a peculiar structure that leaves even strong students confused and disoriented well into their first term.

The JD is a basic bachelor’s degree that is missing Law 101. It is, in effect, three years of the second year of an English, Sociology, or Chemistry degree. We don’t give students at the outset the basic orientation in the discipline they need to make sense of the insider baseball in which most case law is written.

To shift the metaphor, a law degree is really a degree in a second language. By starting students with the standard five substantive courses, we thrust them into advanced Spanish or Russian without first teaching the grammar they need to put sentences together.

But law profs compound the problem by assigning substantial excerpts from dozens of Supreme Court decisions — without the headnotes!

The theory behind this practice is that students must grapple with the jurisprudence itself, or fragments of it, without any tool to help orient them, so as to experience a gradual transformation into effective legal reasoners.

At least that’s the theory. Let’s talk about practice.

The reality is that today’s students belong to a generation in which fewer than 16 per cent read for pleasure each day. In a 2022 survey, fewer than half of American adults had read a single book the previous year. Many readers leaving high school struggle to piece together the argument in a short article. Profs at Ivy League universities report that some students can barely get through a single book in a semester, let alone several.

And so incoming 1Ls are being asked to read 15-20 cases a week, without headnotes and now, in some classrooms, without AI. What really happens?

An alternative to the ‘make them struggle’ approach

What’s missing from discussions of AI use is just how confused and disoriented students are for the first two or three months of law school. Given how much is coming at them, and how fast things are moving, by the time they begin to understand how cases are structured and what judges are saying, the prof has already covered much of a foundational subject, like property, contracts, or constitutional law.

Years ago, I began assigning my 1Ls the headnotes of Supreme Court decisions, often accompanied by a short portion of the reasons. Two or three times a semester, I assign the entire decision. This works well in criminal law, where many of the cases are recent and the headnotes are extensive. Students appreciate having less to read, and I have not noticed a marked difference — positive or negative — in the quality of analysis on the exams.

I believe encouraging students in 1L to use AI — where they think it might help — to create 400- to 600-word case summaries and to query a chatbot about the basic contours of a decision could take this approach further. By using AI to obtain an overview or clarify difficult details, students could find their bearings much sooner.

Used in this way, AI could help students absorb the salient points of the material earlier and improve their comprehension as the year unfolds.

Empirical studies lend support to this.

A study by faculty at the University of Minnesota Law School, published in April 2026, tested AI’s effect on upper-year students. One group used AI for some tasks and a control group did not use it.

One task required students to synthesize legal principles from complex source materials. The authors found that “participants who used AI produced substantially stronger synthesis memos than participants who did not. In fact, AI availability increased performance on the synthesis task by roughly 50 to 70 percent.”

The results on comprehension were more surprising. The authors found “no evidence that the use of AI during the synthesis task degraded performance on either the comprehension or application tasks, even though AI was no longer available to either group for these tasks.”

And when it came to a task that tested student’s ability to apply the knowledge they had synthesized, the authors found “the exact opposite of our hypothesized effect. Participants who used AI to complete the synthesis task outperformed the control group on the application task, even though neither group had access to AI.”

The authors’ best explanation was that AI helped more students produce strong memos, which in turn improved their later reasoning when AI was no longer available. The better your synthesis memo, the better your ability to apply the law.

The study involved upper-year students, who may have benefited from an AI-free 1L. Its findings do not tell us what would happen in first year. But they support the possibility that AI could help more 1Ls gain a basic conceptual grasp of what they read and apply the law more effectively, earlier in the process.

A second study, “Law Professors Prefer AI Over Peer Answers,” looked at how well AI could serve as “effective tutors” in contract law courses. Surveying sixteen profs across the US, the study found that “in a blind evaluation of nearly 3,000 anonymized comparisons, professors rated AI responses significantly higher than answers written by other professors, with AI winning 75% of head-to-head matchups.” The study also found that “LLM responses were also rarely flagged as harmful (3.53% vs 12.06% for professors).”

In short, depending on how it is used (a topic for another post), AI can now provide law students with consistently clear and informative answers.

A tale of two law students and two law schools

I’ll close with a contrast. Consider two 1L students: one uses AI and the other doesn’t.

The student who does not use AI spends much of the first two to three months trying to understand what judges are saying in the dozen or so cases they’re asked to read each week. By the time they begin to find their bearings at the beginning of the second term, the course has moved on. They have little time to make sense of the seminal cases quickly passing by each week.

Meanwhile, the student who uses AI forms a habit, from the outset, of distilling each case into a comprehensible summary of its basic contours and main holdings. They see the bigger picture sooner and more clearly. They have a more rewarding experience; they feel more confident and engaged; and they’re better placed to ask higher-order questions about things like why the law took a certain direction or what policy considerations shaped the decision.

To be clear, I am not suggesting that students use AI instead of doing their assigned reading. I see it being used as part of a broader shift toward more realistic reading demands in 1L. We cover so much ground in so many classes in law school, and far too often we prefer wandering through dense thickets of detail instead of helping students see the forest for the trees.

Some students will find their bearings quickly without AI. But there is reason to think that more students would reach that point sooner with AI. If so, insisting that they struggle without it may do more to hinder their learning than to help it.